Hello, ALL.
I have to diminish the impact on an operations' server reading process, otherwise server may "get stuck" and stop business operations. Even slowing much this server would be unimaginable to business.
I know about Throttle tool, and how it allows users to slow processes speed from that point on.
I would like to ask whether putting the Throttle tool right after an Input Tool connecting to this server would slow server reading and help not consuming much of this server's resources. In other words, will server reading happen at full speed and then slow down when passing data through Throttle tool, or the requests to server will be slower due to this Throttle tool.
I have tried this with an Excel file, it indeed diminishes reading speed. I just would like to confirm whether does this happen in the case of reading a server.
Depending on the answer, I can use Alteryx connecting to this server to allow my internal customer to raise previous month's production numbers without being worried about "taking server down" or even slowering it.
So, what can you tell me about this?
Thanks for your attention,
Bruno (@Brazil)
Solved! Go to Solution.
Hi Bruno,
Putting a Throttle tool after your Input tool will not slow down server reading. It just slows down the speed of the next/downstream tool - https://help.alteryx.com/9.5/Throttle.htm
First thing I would recommend is to not be querying from a production/operations server, and instead get a read replica server setup which is dedicated for reporting queries.
If this is not an option, I can only think of 1 method that may help.
Let's assume that
Start off with a Text Input tool with a list of start and end dates. Or it would probably be easier to create a list of date ranges in an Excel file and use that instead. Or dynamically create a list using tools within Alteryx, (formula tool / generate rows tool)
E.g. if you want to query X days worth of data every X minutes, setup a data source like this:
Then push the list into the Throttle tool and set the number of records (date ranges) you want to process each minute.
Then add a Dynamic Input tool, using the startdate and enddate values to replace some "placeholder" dates that you have entered in your SQL (or equivalent language) query:
Below shows an example of what the workflow would/could look like:
Hope this helps,
ryan